[sdiy] additive synthesis - stretched harmonics

[Richard Wentk]

At 13:46 09/04/2006, Paul Perry wrote:
>I wouldn't expect a violin string to have
>stretched frequencies, unless you played it
>by plucking it.
>The sliding effect of the bow, cannot produce
>overtones that are not exact multiples.

The bow actually produces chaotic/randomised microplucks, which is one 
reason the sound is so interesting - it's not a steady waveform at all, 
there's a lot of internal detail, especially during a real performance 
where the bowing is changing all the time.

Also, because the string isn't perfect there are some non-linearities in 
the response. So again, you wouldn't expect an ideal string model to be 
accurate.

>this is all abit off-topic, except that it does illustrate the
>difficulty of REALLY synthesising a replica of an acoustic
>instrument.... something tht it doesn't hurt to remember, ocassionally.

Indeed.

Richard